Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-52774

Periodically persist metrics about resharding on shards

    • Sharding
    • 2

      The persisted counters will be used to surface estimates about the resharding operation in $currentOp. Persisting and replicating the metrics enables $currentOp to report these estimates following a primary failover or server restart.

      • The metrics from the "resharding" section in server status can be a starting point for the ones to persist. It should also be possible to reinitialize the in-memory counters from the persisted state.
      • The counters for the number of successful, failed, and user-canceled resharding operations need not be persisted. The counters therefore need not be persisted on the config server.

            Assignee:
            backlog-server-sharding [DO NOT USE] Backlog - Sharding Team
            Reporter:
            max.hirschhorn@mongodb.com Max Hirschhorn
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

              Created:
              Updated:
              Resolved: